|
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
|
|
|
|||||||||
|
|||||||||||
|
J Am Coll Cardiol, 2007; 49:2081-2089, doi:10.1016/j.jacc.2006.08.069
(Published online 11 May 2007). © 2007 by the American College of Cardiology Foundation |
,2
* Cardiovascular Biology Research Centre, Division of Cardiac and Vascular Sciences, St. George’s, University of London, London, United Kingdom
Analytical Unit, Division of Cardiac and Vascular Sciences, St. George’s, University of London, London, United Kingdom
University Cardiology Department, Democritus University of Thrace, Alexandroupolis, Greece.
Manuscript received March 14, 2006; revised manuscript received August 14, 2006, accepted August 21, 2006.
* Reprint requests and correspondence: Dr. Juan Carlos Kaski, Cardiovascular Biology Research Centre, Division of Cardiac and Vascular Sciences, St. George’s, University of London, Cranmer Terrace, London SW17 0RE, United Kingdom. (Email: jkaski{at}sgul.ac.uk).
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Background: Plaque rupture in ACS depends at least partly on the volume of the necrotic lipid core. Histopathologic studies have suggested that cholesterol transported by erythrocytes and deposited into the necrotic core of atheromatous plaques contributes to lipid core growth.
Methods: Consecutive angina patients were prospectively assessed; 120 had CSA (83 men, age 64 ± 11 years) and 92 ACS (67 men, 66 ± 11 years). Total cholesterol content in erythrocyte membranes (CEM) was measured using an enzymatic assay, and protein content was assessed by the Bradford method.
Results: The CEM (median and interquartile range) was higher (p < 0.001) in ACS patients (184 µg/mg; range 130.4 to 260.4 µg/mg) compared with CSA patients (81.1 µg/mg; range 53.9 to 109.1 µg/mg) (analysis of covariance). Total plasma cholesterol concentrations did not correlate with CEM levels (r = –0.046, p = 0.628).
Conclusions: This study shows, for the first time, that CEM is significantly higher in patients with ACS compared with CSA patients. These findings suggest a potential role of CEM as a marker of atheromatous plaque growth and vulnerability. Large ad hoc studies are required to establish the clinical importance and pathogenic significance of CEM measurement.
| ||||||||||||||
Aggregation of lipoproteins and their phagocytosis or endocytosis by macrophages contribute to the accumulation of cholesterol within plaques (6) and thus to the growth of the lipid core. However, although apoptotic macrophages are an important source of cholesterol within plaques, it is unlikely that all of the cholesterol contained in plaques derives from foam cells alone. Most of the cholesterol in foam cell is esterified (7), whereas the atherosclerotic lipid core has a remarkably high content of free cholesterol (6).
Arbustini et al. (8), in patients with chronic thromboembolic pulmonary hypertension, and Kolodgie et al. (9), in patients who died suddenly of coronary causes, observed that advanced atherosclerotic plaques contained erythrocyte membranes in the necrotic core, suggesting that red blood cells may actively contribute to plaque growth. It has been shown that the content of free cholesterol in erythrocyte membranes exceeds that of all other cells in the body, with lipids constituting 40% of their total weight (10). Furthermore, recent studies from Kolodgie et al. (9), Purushothaman et al. (11), and Moreno et al. (12) have shown that angiogenesis and intraplaque hemorrhage are important in the development of the vulnerable unstable atherosclerotic plaque.
It has been suggested that cholesterol contained in red blood cell membranes participates in the rapid progression of the atheromatous core and may lead to atheromatous plaque instability. The contribution of cholesterol content of erythrocyte membranes to the clinical presentation of coronary artery disease (CAD) and the rapid progression of coronary atheroma has not been studied. We therefore sought to assess whether total cholesterol content of erythrocyte membranes (CEM) differs between patients with stable and those with unstable CAD and may thus represent both a marker of plaque vulnerability and a pathogenic mechanism of ACS in CAD patients.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Stable CAD was defined as typical exertional chest pain relieved by rest and/or nitrates, without a change in frequency or pattern for 3 months before study entry, and with a positive response (>1 mm ST-segment depression) to exercise stress testing. Myocardial infarction was diagnosed in the presence of prolonged (>20 min) chest pain, with ST-segment changes suggestive of myocardial ischemia or necrosis on the standard 12-lead electrocardiogram, associated with increased serum markers of myocardial damage measured on at least two occasions during the first 24 h after the index event (>2-fold increase over the upper normal range required for creatine kinase-myocardial fraction and troponin T). Myocardial infarction patients were considered to have ST-segment elevation myocardial infarction in the presence of 
0.2 mV ST-segment elevation at the J point in 2 or more contiguous electrocardiogram leads. We diagnosed non–ST-segment elevation myocardial infarction in the presence of new ST-segment depression (0.1 mV) and/or T-wave inversion (0.3 mV) in two or more contiguous leads (13,14). Unstable angina was defined as anginal pain at rest fulfilling Braunwald’s IIIb criteria with transient significant ischemic ST-segment or T-wave changes, or both, without evidence of myocardial damage (15).
The present study did not include patients with a history of a previous ACS; excessive alcohol intake; hematologic, renal, liver, or thyroid diseases; or malignancies. Furthermore, patients with infectious or autoimmune diseases, familial hyperlipidemia, and those undergoing surgical procedures in the preceding 3 months were excluded from the study. None of the patients in the study was receiving treatment with anti-inflammatory drugs or hormone replacement therapy. We also excluded patients with abnormal red blood cell (RBC) counts (<4.7 and >5.9 x 106/µl for men and <4.2 and >5.4 x 106/µl for women) and/or abnormal hemoglobin levels (<13.5 g/dl and >18 g/dl for men and <12.5 g/dl and >16 g/dl for women). The study was approved by the local research ethics committee, and all patients gave written informed consent before study entry.
The patients’ baseline clinical characteristics at initial presentation are summarized in Table 1.
|
75% reduction in lumen diameter (14). Stenosis score was used to assess the severity of CAD on angiography. Stenosis score, as described in previous studies from our group (16), is aimed at reflecting the most severe stenosis observed in each of the main coronary vessels assessed. Interobserver variability for vessel score was 3.8%. Laboratory analysis. In every patient, peripheral blood samples for CEM and other biochemical variables were obtained after a 12-h overnight fast, at the time of coronary angiography immediately after the catheterization of the femoral artery, and before the infusion of heparin and angiographic contrast medium. Blood specimens for CEM analysis were collected in standard vacutainer tubes containing citrate. A 3-ml aliquot of venous blood collected in citrate plasma tubes was centrifuged at 1,500 rpm for 10 min at 4°C; the plasma and buffy coat were carefully removed by aspiration, and the remaining RBCs were resuspended and washed twice in 154 mM NaCl isotonic solution (17). One milliliter of washed erythrocytes solution was hypotonically lysed in 30 volumes of cold distilled water, mixed by vortex, and allowed to stand for 15 min. Membranes were separated from the hemolysate (supernatant from hemolyzed RBCs) by centrifugation at 15,000 rpm for 15 min at 4°C, this step was repeated 3 times until a white/pale pink pellet containing hemoglobin-free erythrocytes (ghosts) was obtained (18).
Erythrocyte ghosts were resuspended in 1 ml of PBS and stored at –20°C until further analysis. A 125 µl aliquot of membrane suspension was used to determine membrane protein concentration by the method of Bradford using bovine serum albumin as a standard (19). Red blood cell membrane lipid extraction was carried out from a second 500 µl aliquot following Folch’s method (20).
Total cholesterol was measured on RBC membrane lipid extracts by means of a commercial enzymatic assay (Waco Chemicals GmbH, Neuss, Germany) (21) following the instructions of the manufacturer. The detection limit of the assay was 1.8 mg/dl, and the manufacturer’s reported intra- and interassay precision were both <1.1%. Briefly, a 6-point calibration curve was prepared by diluting the standard solution provided in the kit, the absorbance of each sample was measured against blank at 505 nm and the result was plotted against the calibration curve to obtain the amount of total cholesterol. All the samples were measured in duplicates and none of the duplicates had a coefficient of variation higher than 4%. We were able to measure the amount of cholesterol in all the samples assayed. Results are expressed as micrograms of total membrane cholesterol per milligram of membrane protein.
C-reactive protein (CRP) was measured on the COBAS Integra (Roche Diagnostics Limited, Lewes, East Sussex, United Kingdom) using the CRP latex assay in both the high-sensitivity application (analytical range 0.2 to 12 mg/l) and the normal application (analytical range 2 to 160 mg/l). All other biochemistry measurements were carried out using standard methods.
Statistical analysis. Results for continuous variables are presented as means and SD or as medians and interquartile ranges if the distributions were skewed and as percentages for categorical data. The 2-tailed unpaired Student t test or the Mann-Whitney U test were used to evaluate differences in continuous variables between the 2 groups. Comparisons between categorical variables were performed with the chi-square test or Fisher exact test as appropriate. Normality was tested using the Kolmogorov-Smirnov test. The CEM, CRP, creatinine, triglycerides, total cholesterol, low-density lipoprotein (LDL) cholesterol, and high-density lipoprotein (HDL) cholesterol levels were not normally distributed and were therefore logarithmically transformed as required to approach normal distribution and obtain equal variances. Analysis of variance with covariates (ANCOVA) was used to evaluate differences in CEM levels between stable CAD and ACS patients after adjustment for all the variables that were significantly different between the two patient groups.
Because preliminary analysis suggested that association between CEM levels and CAD instability was more of a threshold effect, CEM levels were assessed as a dichotomous variable using its median value as a cutoff point. Similarly, continuous variables with skewed distribution were also assessed as dichotomous variables using as cutoff points their risk-associated levels in CAD patients (i.e., total cholesterol 4.5 mmol/l; LDL 2.5 mmol/l, and HDL cholesterol 1.1 mmol/l; triglycerides 1.7 mmol/l; CRP 3 mg/l, and creatinine 107 mmol/l) (22).
Simple logistic regression analysis was used to assess univariate associations between patients’ characteristics and biochemical measurements and CAD status (stable CAD vs. ACS). Multiple logistic regression analysis was used to assess the independent adjusted relationship between CEM levels and CAD status, with independent variables being those with p < 0.05 on univariate analysis. Odds ratios (OR) with 95% confidence intervals (CI) were calculated for all patients (n = 212, stable CAD and ACS). The OR for dichotomized continuous variables represents the relative risk between levels above and below the prespecified cutoff point. The OR for categorical variables represents the relative risk between the presence or absence of the variable. Correlation analysis between variables was carried out by Spearman’s correlation coefficient (r). Receiver-operating characteristic (ROC) curves were calculated for CEM, 1/HDL, and CRP levels. A p value <0.05 was considered to indicate statistical significance. The SPSS 11.0 statistical software package (SPSS Inc., Chicago, Illinois) was used for all calculations.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Total cholesterol content of erythrocyte membranes. The CEM was significantly higher (p < 0.001) in the unstable CAD patient group (184 µg/mg, range 130.4 to 260.4 µg/mg) compared with patients with stable CAD (81.1 µg/mg, range 53.9 to 109.1 µg/mg) (Fig. 1). Analysis of covariance showed that CEM remained significantly higher (p < 0.001) in the unstable CAD group after adjustment for all the variables that were significantly different between the 2 groups (R2 = 0.437; for full model p < 0.001). In detail, estimated covariate-adjusted means with 95% CIs for CEM levels among the 2 study groups assessed with the ANCOVA model were 80 µg/mg 95% CI 71.9 to 89.1 for CSA patients and 181.1 µg/mg 95% CI 159.6 to 206.1 for ACS patients.
|
The proportion of patients with ACS across CEM deciles (Fig. 2) shows that the association between CEM and CAD status appears to be more of a threshold effect. In view of this finding, we used the median CEM level (112.1 µg/mg) as a cutpoint and conducted evaluations for evidence of association between CEM and CAD status (stable vs. ACS) above and below the aforementioned cutoff value.
|
|
|
In addition, high CEM levels were associated with ACS status both in a subgroup of patients with high total cholesterol levels (n = 125, 4.5 mmol/l) (OR 35.54, 95% CI 12.02 to 105.2; p < 0.001) and in a subgroup of patients with low cholesterol levels (n = 87, <4.5 mmol/l) (OR 29.47, 95% CI 9 to 96.47, p < 0.001).
The use of statins was associated with lower CEM values (n = 100, 84.8 µg/mg, range 53.9 to 135.7 µg/mg) (p < 0.001) compared with patients not receiving lipid-lowering treatments (n = 112, 147.2 µg/mg, range 106.6 to 212.6 µg/mg). In view of the association observed between statin use and CEM levels, we compared the predictive ability of CEM in both ACS patients with and without statin treatment. In the subgroup of patients receiving statin treatment, CEM levels were predictive of ACS (OR low vs. high 30.6, 95% CI 8.59 to 109, p < 0.001). Furthermore, CEM levels were also associated with ACS (OR low vs. high 21.33, 95% CI 7.43 to 61.22, p < 0.001) in the subgroup of patients not receiving statin treatment.
As disrupted or ulcerated atherosclerotic plaques are associated with ACS (23) and rapid disease progression (24), and often appear as "complex" stenoses at angiography, we assessed lesion morphology in all stenoses with 75% diameter reduction. Briefly, stenoses were subdivided into complex or smooth, as reported in previous studies by our group and others (23,24), and the presence or absence of complex angiographic lesion morphology was correlated with CEM results. We observed that the presence of angiographically complex coronary lesions was associated with a significantly (p = 0.027) higher CEM (110.5 µg/mg, range 71 to 183.7 µg/mg) compared to absence of such morphology in coronary angiography (92.9 µg/mg, range 53.9 to 111.5 µg/mg). Furthermore, CEM levels were also significantly positively correlated with the number of complex lesions (r = 0.272; p = 0.003). However, CEM levels were not correlated with number of diseased vessels (r = –0.080, p = 0.246) or with stenosis score (r = 0.024, p = 0.730) observed on coronary angiography.
ROC analysis. Consistent with these results, ROC analysis regarding predictive accuracy for patients’ CAD status showed that for CEM, the area under the curve was 0.885 (0.840 to 0.931); p < 0.001, for CRP 0.737 (0.669 to 0.806); p < 0.001, and for 1/HDL cholesterol 0.711 (0.640 to 0.781); p < 0.001. Receiver-operating characteristics analysis, albeit not controlling for possible confounders, indicates that CEM was a better marker of CAD activity than either CRP or HDL cholesterol levels (Fig. 3).
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
These observations are in keeping with the results of histopathology studies suggesting that the instability of atherosclerotic plaques may be mediated at least in part by erythrocyte membrane cholesterol (9) through an enlargement of the lipid core. Arbustini et al. (8) reported that atherosclerotic plaques outside the coronary circulation contain glycophorin, an exclusive component of erythrocyte membranes (29). Consistent with the findings of Arbustini et al. (8), Kolodgie et al. (9) found erythrocyte membranes to be present in the necrotic core of advanced coronary atheroma. Moreover, they also showed that erythrocyte membrane cholesterol contributes to an abrupt increase in the content of cholesterol in the necrotic core of coronary lesions. Thus, taking these findings together, it is conceivable that CEM may contribute to the expansion of the necrotic core and, in turn, to plaque instability. Various investigators (9,11,12,30) have reported that intraplaque hemorrhage and plaque neovascularization are associated with plaque vulnerability and the rupture of the fibrous cap of atheromatous lesions.
Although in the present study we did not specifically investigate plaque instability, our findings suggest a possible link between CEM levels and clinical instability, which may, in turn, reflect atheromatous plaque instability. This issue, however, requires further investigation. The potential mechanisms whereby erythrocyte membrane derived cholesterol can contribute to plaque instability are speculative at present but may include: 1) an abrupt expansion of the necrotic core (3–5); 2) a change in the lipid composition of the atheromatous core favoring a predominance of free cholesterol (31,32); 3) inflammatory mechanisms, because erythrocyte membranes can bind a wide array of chemokines such as interleukin-8 (33); 4) erythrophagocytosis that may lead to foam cell formation and macrophage activation (34); and 5) a combination of these mechanisms.
Lipid contained in erythrocyte membranes results from an exchange of lipoproteins between erythrocytes and the circulation (35), as erythrocytes are not capable of synthesizing lipids de novo. In our study, however, circulating cholesterol levels did not correlate with CEM content, probably because CEM content represents the equilibrium achieved between red cell cholesterol influx and efflux, and intracellular cholesterol transport. Several enzymatic and protein-mediated pathways are involved in the handling of cholesterol by the erythrocyte, which are regulated by complex feedback mechanisms and genetic factors that regulate protein expression and enzymatic activity (36). In contrast to our results, studies (37,38) have reported an association between plasma cholesterol and CEM levels. An explanation of this disparity may be that the aforementioned studies showed that a reduction in plasma lipid levels was associated with a similar reduction in CEM levels. However, none of the studies reported a direct correlation between CEM levels and plasma lipid concentrations as assessed by specific correlation statistical tests. The mechanisms responsible for increased CEM in patients with ACS require investigation. Increased cholesterol uptake by the erythrocyte membrane, reduced cholesterol efflux, or both may be responsible. The relationship between CEM and circulating cholesterol levels also deserves further investigation.
The CEM levels were lower in patients receiving treatment with statins compared to patients without lipid-lowering treatment, suggesting that the exchange mechanisms responsible for cholesterol uptake and/or efflux by the erythrocyte can be affected by statins. This is in agreement with previous studies that have reported that statin treatment decreases cholesterol content of erythrocyte membranes in patients with hypercholesterolemia (37,38). Although the use of statins could beneficially affect erythrocyte membrane lipids, the association between increased CEM and ACS in the present study was independent of lipid-lowering treatment. The intriguing finding in our study that statin use is associated with CEM levels, whereas cholesterol levels are not, suggests that CEM levels are independent of circulating cholesterol levels, and statin treatment may have an effect that does not necessarily parallel that on cholesterol plasma levels. Of importance, other studies (39) reported that diet-induced reductions in circulating lipid levels were not paralleled by reductions in CEM levels, further strengthening our hypothesis that the effect of statins on CEM levels may be at least partially independent from the reduction they cause in plasma cholesterol concentrations.
Although our study was not aimed at assessing the association between CEM levels and angiographic CAD, we observed that CEM levels were associated with the presence of angiographically complex coronary lesions, a predictor of rapid disease progression and adverse prognosis in established CAD patients (23,24). These results endorse the hypothesis that increased levels of CEM may favor plaque instability and the development of ACS. Our finding that CEM levels were not associated with extent or severity of CAD is in agreement with previous studies (40) that suggested a lack of association between the extent and severity of coronary stenosis and the risk of myocardial infarction and unstable angina.
In summary, the present study has shown for the first time a significant independent association between CEM in circulating red cells and CAD clinical presentation. Our results are novel and intriguing and indicate not only that CEM may be a marker of clinical instability, but also a promoter of plaque instability and atheromatous core growth. The findings of the present study should be interpreted in light of certain limitations. First, our investigation represents a small observational study; however, it was powered enough to detect differences in CEM levels among study groups. Even with the relatively small sample size, the present study is hypothesis-generating. Larger studies are required to investigate further the predictive value of CEM. Second, underlying pathogenetic mechanisms are still speculative. Our results, together with the evidence of presence of erythrocytes in ruptured atherosclerotic plaques documented in previous studies (8,9), have important pathophysiologic implications, and the correlation between CEM and the clinical presentation further links the erythrocytes with pathogenetic events. However, it is not known how and by which mechanism erythrocytes become loaded with cholesterol, or to what extent plaque neovascularization and plaque positive remodeling contribute to CEM-related risk. Finally, we did not determine blood viscosity, a factor that could increase the risk of ischemia, especially in CAD patients (41). However, it is unlikely that a difference existed in blood viscosity among ACS and CSA patients, as there were no differences in the major determinants of blood viscosity (i.e., hemoglobin levels and red blood cell count) between the 2 groups (42). Differences in CEM levels could not account for a significant difference in viscosity because the impact of lipid composition of erythrocyte membranes on blood viscosity is minimal (38). Furthermore, the clinical relevance of an association between these rheologic factors and the risk of coronary heart disease remains unclear (43). Altered lipid composition of the erythrocyte membranes may affect platelet reactivity and thrombus formation and could represent a pathogenic link between CEM and ACS (44). However, the clinical relevance of such a relationship requires investigation (44). Nevertheless, our findings indicate that ACS patients have raised CEM levels, and these levels cannot be explained simply by the presence of underlying risk factors, clinical features, and concurrent medications. Further studies are warranted to elucidate the role of cholesterol contained in RBC membranes as both a marker of plaque instability and a pathogenic mechanism of rapid CAD progression.
|
Footnotes |
|---|
2 Dr. Romero is currently affiliated with the Lipid Laboratory, Clinical Chemistry Department, Gregorio Marañón University Hospital, Madrid, Spain.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Related articles in JACC:
This article has been cited by other articles:
|
|
 |
|
  J. A. Ambrose In Search of the "vulnerable plaque": can it be localized and will focal regional therapy ever be an option for cardiac prevention? J. Am. Coll. Cardiol., April 22, 2008; 51(16): 1539 - 1542. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Herrmann and A. Lerman Atherosclerosis in the Back Yard J. Am. Coll. Cardiol., May 29, 2007; 49(21): 2102 - 2104. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Arbustini Total Erythrocyte Membrane Cholesterol: An Innocent New Marker or an Active Player in Acute Coronary Syndromes? J. Am. Coll. Cardiol., May 29, 2007; 49(21): 2090 - 2092. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
